Skip to content

setup.md

Latest commit

 

History

History
85 lines (64 loc) · 1.74 KB

setup.md

File metadata and controls

85 lines (64 loc) · 1.74 KB

Setup Guide

1. Clone the Repository

git clone https://github.com/embed2scale/AI-for-Good-Tutorial-2026.git
cd AI-for-Good-Tutorial-2026

2. Create a Virtual Environment and Install Dependencies (uv or pip)

python3.12 -m venv .venv
source .venv/bin/activate
pip install --upgrade pip
pip install git+https://github.com/terrastackai/terratorch.git

git lfs install --skip-smudge
git clone https://github.com/embed2scale/NeuCo-Bench.git

pip install -r requirements.txt

Note: If TerraTorch fails to install because of GDAL or other geospatial libs, continue with Section 3 and retry.

3. Install GDAL (If Needed)

GDAL is required for geospatial I/O. Skip if TerraTorch installed without errors.

macOS (Homebrew)

brew install gdal
pip install gdal

Ubuntu / Debian

sudo apt-get update
sudo apt-get install -y gdal-bin libgdal-dev
pip install gdal

On other systems try

pip install gdal

Then retry:

pip install git+https://github.com/terrastackai/terratorch.git

4. Download Data

Download labels

python scripts/download_ssl4eo_downstream.py --out data/neuco_ssl4eo_downstream --labels-only

Download imagery (S2 L1C) — Optional (~50 GB)

Only needed for full end‑to‑end workflows (Part 3).

python scripts/download_ssl4eo_downstream.py --out data/neuco_ssl4eo_downstream --modalities s2l1c

Notebook Usage

We will launch Jupyter via:

jupyter notebook

If you run into kernel issues, register a clean one:

python -m ipykernel install --user --name tutorial-kernel --display-name "Tutorial Kernel"

Next Steps

You're fully set up. Head back to the main README to start the tutorial.